yeah that's about the point where loss usually peters out

Why I can't use "Import files"

Ah...Why...

The Wavs Folder Needs To Be Named Ufeffwavs

Not, it doesn't

Format your transcription as utf8 without bom

This solves the problem

Still working on this but I’m having some type of error with the second text what an I doing wrong

Error 1: it expect the file to be

.wav

, not

.wavs

Also, use spaces instead of underscores

Also the Nb is outdated, use the ones in #841437191073955920

I also have it in like 22K hertz

also the transcript is supposed to be in notepad not colab

Why the heck are you using a notebook that's clearly labled as outdated?

I’ve never actually like train an AI and I’ve heard that a lot of people use a notebook for it

Okay, but that doesn't answer my question. You saw that the notebook was outdated, so why did you try to use it?

I thought it would work what is the right formant notebook

(((#841437191073955920)))

Unless you want legacy specifically then https://colab.research.google.com/drive/1OFhKJF846EbFsvdNVkapSZvXc7wHvoSd

What are you doing?

My friend is having this error using the legacy Tacotron 2 training notebook: @Cris140 Can you help me to solve that?

Which notebook is he using?

The legacy TT2 training notebook

to me it looks like they didn't run the previous cells

Probably that

With any problem, it's always best to retry and check if the problem persists, sometimes for unknown reasons the git clone code doesn't work, so that could be the case

I’m still like trying to figure this out

1 __author__ = 'tsungyi'
      2 
----> 3 import pycocotools._mask as _mask
      4 
      5 # Interface for manipulating masks stored in RLE format.

pycocotools/_mask.pyx in init pycocotools._mask()

ValueError: numpy.ndarray size changed, may indicate binary incompatibility. Expected 88 from C header, got 80 from PyObject

got this error for controllable talknet colab